Northern's Sterlingsilber M52 Turbo 325is

[Northern]

Not much going on with this car. Drove it a bit but I have no reason to leave the house.

Still trying to get the MAP working.
wideband O2 sensor died again, so I warrantied it but have yet to install the new one.


Pulled out Ye Olde Fender Roller to try to lower the car down again. I don't know if it made much of a difference, but it let me lower the car an inch or so.



Also lets me take pictures at utter degenerate levels of stance



I bought this kit for the S2.93 I'm building. I honestly though there would be more to it for $100. Would not recommend, just buy everything separately.
Hell, you don't even need the axle clips and I don't know why there are two cover gaskets.



MAP progression. Bought a third sensor, AC Delco 4bar for some LS variant. It also has a bosch part number.
It's supposed to mount to a flange, so I had to adapt it for a hose barb.





Much smaller than the old one. Probably will reuse the old one on the other car, since I think the issue is from trying to use the post cat O2 ground.



Really want to be able to log MAP because my EBC just flashes "CUT" over 10psi and it seems that the 7.5psi spring wants to live somewhere north of 10psi.

I reenabled VANOS and load comes up faster. I have the modifier tables changing fueling/timing much less than stock.

Engine seems happy with a bottle of VP Octanium per 35L of fuel. Hitting a hair above double the stock load numbers (since the MAF and Load in this graph are half the true number), but since this is a different MAF, something could be off.

[Northern]

Might need to open this pic in a new tab for it to show a decent resolution...

[Northern]

Small update just to remember what I did from Jan-Feb

Replaced my Wideband O2 sensor again (warranty). It's an NTK Xref of a Bosch LSU 4.2.
Apparently that's what my innovate LC-1 shipped with in the first place, so I don't feel sketchy about running it.
The fact that I've burned through 4 of them in 2020 is probably a sign that I need to delay the O2 from starting until the car is actually running.


I finally got my MAP sensor working, turns out there's something with the rear O2 sensor's sensor ground in the ECU that stops it from being used for anything else, so I spliced into one of the large ground splices in the engine bay:

My Revised setup is:

+5V from Tank Pressure Sensor (X8000 Splices into TPS +5V fed from the ECU X6000-44)
Signal Input to Post Cat O2 #1 Signal IN (X6000-77)
GND to BR/OR Chassis GND Engine Harness Splice (X6459)

X6459 is the large splice of Brown/Orange GND in the engine bay.
No reason I chose this one in particular, other than that I found it quickly visually and in the ETM, and confirmed it went to the strut tower GND on the RHS of the car.

I Noticed it read ~6kPa high at atmospheric, so I used my boost leak tester to pressurize the intake and hoped to use my EBC to get a reading.

I didn't think far enough ahead because as soon as my EBC saw positive pressure it went straight to flashing "LEAN" since the car wasn't running,
then when I passed the 10.5psi boost cut, it started flashing "CUT" at the same time as "LEAN".
After I lowered the pressure, the gauge permanently read "17.5" and after calibrating the MAP I thought I had somehow broken it until Harrison from GFB told me that I had just switched the readout to AFR mode by holding the settings button as I freaked out the lean message.

Anyway, I used my analog boost gauge to correlate with voltage in my romraider log. Despite how sketchy it felt, it worked to make an adjusted scalar that looks correct.


Then I spent 3 attempts at fixing a fuel leak at the fuel rail.

I thought it was something wrong with my 3d printed spacer arrangement, since the fuel rail didn't look to be sitting straight, maybe the spacers were too long (they were, and we also crushed).
I replaced the spacers with shorter/unsquashed ones, and the rail still leaked.
I replaced these spacers with harder HDPE ones, and the rail still leaked.
Then I replaced the top Orings on the EV14s, and the rail still leaked.
Then I noticed the top of the rail was wet when I was priming the system. I just happened to look under the rail and saw:

[moatilliatta]

Was the fuel leak from the spot weld or corrosion?

Hope you can get the car sorted out soon. Have you though about just going standalone if you're trying to use a map sensor?


That turnout quite well! I may have to do the same, I always hated how heavy that panel was. Some cars seem to have steel or aluminum heat center heat shield. I think it maybe and early early car vs late


Hah, this is great. A local at the time, I helped him build the car in his garage. The gold old days..

[Northern]

It looks like it was a spot of corrosion. I couldn't find another early e39 M52 rail, so I ordered some injector hats and a clean looking OBD1 rail from socal, and then use my pitted OBD1 rail on my 4dr.

The standalone question is a long one...
I've thought about buying a Link E36X a few times, (previously about an ECUmaster EMU Black + EDL-1 logger/bt unit) but the E36X is cheaper (and newer) by a bit.
The PNP aspect is my least favorite part of the E36X, because it's OBD1 pinout, but also because adding/removing pins is painful vs say an AMP Superseal motorsport connector on a MoTeC or Haltech.
It would give me live tuning, failsafes, and let me run an e-throttle, traction control, closed loop boost control, etc.

I haven't considered anything cheaper because I don't think it's worth it. Any ECU with poor or non-existent knock control isn't for me, Speeduino and a lot of stuff like lower tier Haltech/Link doesn't do knock, and the M52 on stock CR and 91 is so knock limited on boost that I pretty much have stock timing up until boost hits and then I've fine tuned it using logs of MS41's knock retard and some MLVHD tables.

I think I kind of have the learning curve out of the way with MS41, all the deletes I need are done, MAF hack+ HPX MAF let me deal with the extra air, spreadsheets and MegalogviewerHD views give me the info I need to do quick revisions. The only painful part now is physically removing the ECU to bench flash it instead of live tune. I don't think I have any plans to ever make more than 500whp on this chassis, and even 400+ is unlikely so this is probably how it will stay.

I have probably spent the same amount as an E36X costs, between software, MAF, MAP, EBC, other loggers, so I probably would more seriously consider a standalone if I did it all again and had today's options available. Might still consider it for my 4door if/when I turbo it.

Another big part of it is being part of the MS41.3 beta ( https://sites.google.com/site/openms...-code---ms41-3 ) which has been interesting. The goal is to let this be a public thing once everything has been confirmed to work and there isn't a big chance of adding things (for example, the items in yellow are new since the version I'm running, so that removes the MAF hack and dividing all tables with load by 2 so the variable doesn't overflow, which always raised the question of "what tables have we not found in the ECU that use load")


That exhaust heat shield, I just bent that piece up from a masking tape template of the original, had to make some compromises on shape since I was using a brake for the bends.

Funny to see that engine bay again, I'm probably going to eventually rebuild that turbo (again) for my other car.

[moatilliatta]

Are there any tuning shops local? I'd find just running something the local tuners know by the back of their hand saves a lot of headaches. Haltech seems best bang for the buck with a lot of local support.

Its cool to break into a factory DME, but at what cost.. Plus just nice to be able to log for debugging. I got out of those headaches, I mean cool when it all works. I guess a lot has changed in the last 5 years

Ill go Life Ecu in mine and just run an autosport on the firewall, then build an "Adapter" harness to the ECU, But the Deutsch HD connectors look good too.

Did you put in a baffle into your rear sump setup?

I have the S54 Arc Asylum pan setup on my US S50, I had no idea Tyler did standard E36 pan mod with that subframe.


Also, I have the 2 thick tabs off the subframe, I haven connected a bar yet, looks like id break them... Have you ran a big bar on your with much time on it?

[Northern]

There are like 3 or 4 local dyno shops, but I'd want to tune it myself either way

The logging isn't bad with Romraider Logger, save three things:
You need to drag a laptop with you to datalog
It caps out at like 12Hz vs 100Hz on most standalones, or even 1000Hz on higher end or Link E36X.
MS41 doesn't have MAP, AFR, etc, so you have to either do without, feed them into your logging software (pain in the ass), or wire them into unused inputs in the ECU (there are only 4-5)
No onboard logging - Trying to rectify this with a logger.s module ( https://www.clown-shoe.com/single-po...-data-solution ) but mine is DOA and I've been slacking on returning it to the UK to get it replaced. Added bonus is that this gives 4 more 0-5V inputs to insert in the datalog.

Alternatively, there's an android app called ALDLdroid that lets you log via INPA USB cable on your phone, ba114 more or less pioneered getting that app working with MS41. Only issue with it for me is that the custom MS41.3 flash needs a different calibration file. I'm told Logger.S can output to ALDLdroid as well, but haven't verified this.
Here's an example of how it looks:

https://youtu.be/aIz_ymjrA3s

I didn't baffle the rear sump, just used the cut up stock baffle that came with the modded pan. I fill oil up to the dot above the full mark, but I do get typical lifter tick in autox so I don't think it helps as much as I'd like. Wish I did an S54 sump, really.

I drove the car to work for a year after the swap, did maybe 3-4 autox days, and have maybe put 10k (km) on the car with the IE 25mm bar. Haven't had any cracks yet, but have no idea how badly it flexes so I definitely want to weld a tab on each side next time I have a reason for the subframe to be off.

[Northern]

The whole MAP sensor ground issue was just weird because a lot of people reported that it worked for them, and only when I mentioned that mine didn't work did a few others start poking out of the woodwork with the same issue

[Northern]

Small update.

Same 52lb injectors, new fuel rail setup:
Rust free M50 rail from California
Injector hats from ebay, (clips included, also adds a filter)
Gates 225psi hose







Here's the stupid pit that caused a month of pain swapping O-rings and fuel rail adapters around, then like $300 in parts to swap rails.
Maybe I'll weld up the pit to use this rail again, maybe I won't.




Copied Panici and desoldered/extended the 7segment from the Innovate LC-1.
( I gave up on the APSX G1 since it needed a serial input that is different than their documentation and they weren't willing to help.)









Fingers crossed that it all works.
I originally pulled it to change the SI batteries because the tach kept dying, and hopefully fixed the coolant temp wiggle by cleaning the ground.

(Edit from the future: I went for an hour drive today and it all seems to work. All the issues mentioned are gone so far. The AFR gauge also seems to be happy in its new home.)


While I was messing with the cluster, my clear windscreen for the CBR showed up (a full month earlier than expected too) so I threw that on and now I'm just waiting on a new rear sprocket.

[Northern]

Dash seems to be completely working now, no tach dropout or jumpy coolant gauge.
I also really like having the AFR in the cluster like this.

I've been trying scrape together a 2.93 LSD forever. I think it's 6 years since the swap now.
I have gathered:

e36 328i Ring & Pinion
e30 s3.73 case & LSD guts (could interchange LSD guts/spider gears with the 2.93 but the carrier and cap are a different offset)
s2.79 LSD carrier + Ring & Pinion
Tacingdiffs cap + 4 clutch kit
Timken Pinion Bearings
SKF Carrier bearings
BMW diff seal kit
2 pinion crush sleeves

I started to take the e36 gears out to more or less build a diff from scratch, but I realized there are a few problems with my plan:
The "clean" e30 Diff case I have is rusty on the inside
Most likely going to need to mess with pinion shims, which means installing/removing the pinion a few times, probably ordering a few sets of shims and more crush sleeves.
Sure installing/removing the pinion is *possible* without a press, but it involves heat and BFH so I imagine the bearing wouldn't enjoy that very much

I also have an e30 2.93 open diff that wouldn't need nearly as much work since the pinion is already installed and all.
Case is grubby, but I popped it apart and it was alright inside.

The e30/e36 final drive TIS is out there on the internet, so I just followed those instructions, although the LSD stackup instructions are a little murky and it's lacking some torque specs.
Other than some confusion about how they chose to say some weird shit instead of "convex up" with their Belleville washer instructions, it was smooth sailing to install the racing diffs clutch/plate upgrade and their carrier cap.

FWIW, the spacer plate (or "spacer plates" in this diff, likely because the discs were robbed for my current diff and it looks like they fired a )
I didn't have any clearance issues like some people so I didn't need to machine the cap.

Unfortunately Rockauto has some wrong bearings listed (possibly for 168mm diff?) that of course I bought, so now I'm waiting on the correct ones.
The correct carrier bearings and races for an E30 188mm medium case are:

TIMKEN	LM503349A	(2)
TIMKEN	LM503310	(2)

The ones that came off the diff were SKF bearings, just for anyone wondering.

The Pinion bearings/Races I ordered are:

TIMKEN	HM88510	1
TIMKEN	HM89410	1
TIMKEN	HM88542	1
TIMKEN	HM89449

Seem to be correct just based on part number. What came out were FAG bearings.

Here is the s2.79 (missing clutches/plates) that I need to rob the LSD Carrier from:


Popped the tone wheel off with a chisel.
Give it a whack, rotate, repeat until it starts to come off, then just pry and it pops off



Bottom out a brass drift in the bolt holes and you can give it a few taps and it should start coming loose.



Tada


Same thing with the E30 open 2.93 (Probably will use this someday, so I'm wrapping both gearsets up for another time.


Comparison of 188mm diff pinions: 2.79 vs a 3.73. Huge size (and strength) difference between the two.




Random disassembly + 3 hours of wire-wheeling + rust converter + paint:






Yeah I just brushed on two thick coats of Tremclad. We'll see how that works, my guess is not well, but better than if I had tried to spray it as quickly as I brushed it.

Since the s2.79 had been cannibalized for another 4-clutch diff, I needed clutches/dog ear plates.
On a normal e30 LSD, you would remove the spacer plate to fit a third clutch/plate, and machine the carrier cap by ~2mm to fit the 4th clutch.
Since I needed to buy the clutches/plates anyway, it wasn't that much more to tack on a billet carrier cap to my racing diffs order.

pazi88 runs this exact same setup, and I think we bought our setups only a month or two apart.
When he installed his, he found that his carrier cap needed more machined off it to hit the 0.1-0.4mm spec in the TIS.
Racingdiffs said they corrected the issue after he and others ran into this problem, but I was convinced I would have the same issue.
Turns out mine was fine, with just a hair under the 0.4mm clearance spec.

One thing to note about the racingdiffs set is that I needed to file one spot on the inner spline of the clutch discs. If you try to place it over the spline on the ramp piece, it catches on a burr/defect from how they're cut.



ignoring the tiny washers, my plate stackup (in the orientation they are dropped into the carrier) was:
Bellville Washer (Convex Up)
Dog plate
Clutch disc
Dog plate
Clutch disc
[spider gear/ramp assembly]
Clutch disc
Dog plate
Clutch disc
Dog plate
Bellville washer (Convex Down)


I'm still cleaning up this case. Need to install the ring gear/tone ring/bearings and mess with the shims for lash.

[Northern]

Also, disregard the random chunks of motorcycle clutch in 2 pics of that post...

Kind of an off-topic post:

CBR is fully fixed. Chain, spockets, Stainless front brake lines, Front master cyl rebuild, and new battery (well, I still have new bar grips to install... but don't worry about it)

I didn't ride it this year yet, but it started up nicely.
Video (I think it only plays on flickr. R3v handles this weird so I guess Right click and open in new tab):



My friend Iain was storing a CBR929RR at my place for the winter, we started picking at it.
It had a few issues, but we cleaned the carbs, bypassed the broken knockoff fuel pump/fuel filter and put fresh gas in it, so it does have gas, just no spark.
Coils are getting +12V but there's still no spark. Either something wrong with both coils, the CDI brain, or the VR crank sensor.
Honestly this bike blows my mind because it somehow all just runs from a crank sensor. No cam, TPS, or anything else.




I bought a Link E36X G4X from a member of this forum.
This will not go on this car, since I'm very happy with how MS41 is performing with the 41.3 beta flash, and I already have an EBC for boost control
I don't know if it will go on my 4-dr car this year, because I want to test Alpha-N, Speed Density emulation, and a bunch of other features of the MS41.3 flash, but I will find somewhere to use it eventually.



Also because I've found myself bringing this up a few times recently, here's a pic of a 413 DME (Top) vs a MS41 (Bottom) to show the difference in the connector.



Lastly a teaser pic.

[Northern]

s2.93 LSD went together fine, didn't check lockup because I don't have the tools, and what am I really going to do if it is off anyway.
Can't really check backlash or carrier bearing drag so I yolo'd both of them after a few hours messing with shims.

I'm fairly concerned about the condition of the pinion bearings because they don't feel the best, but I was too lazy to do anything about them so I slammed it all in the car and I'm hoping for the best. Should be very obvious if they're pooched with the solid subframe/diff bushings.

While the diff was off, I figured I was already halfway to having access to the brake hoses above the subframe, so I swapped the original hoses for the BMA Stainless hoses I bought a few years back. Absolute cock of a job, no idea why BMW didn't just put them like 0.5" further forward...


Tuning escapades:

Car looked like it was running like 1 pt richer than last year, but it was just the LC-1 losing calibration. Leaning toward replacement this/next year with something newer.

Trying to resolve a few weird issues:

Car might have weird TPS adaptations (12.53 - 13.06% is where it seems to stabilize - I have seen this range of adapts in logs from others though?)
Car idles ~2000 RPM on start unless I tap the gas pedal, or TPS adapts are reset.
Car goes quite lean on tip-in. - I've tried to add fuel to the main fuel table as a first step.

Possible causes:
M50 Throttle body instead of M52 (maybe TPS is angled differently on the shaft?) - Currently cleaning up an M52 TB and have an OBD2 throttle cable on order to test.
S52 Enrichment tables instead of M52, since they are quite different. - 100% an oversight, probably from when I switched to the S52 software years back.

My main theory about this is that something has to work because turbo/maf/modded software aside, nothing has changed off boost.

[jsta1109]

s2.93 LSD went together fine, didn't check lockup because I don't have the tools, and what am I really going to do if it is off anyway.
Can't really check backlash or carrier bearing drag so I yolo'd both of them after a few hours messing with shims.

I'm fairly concerned about the condition of the pinion bearings because they don't feel the best, but I was too lazy to do anything about them so I slammed it all in the car and I'm hoping for the best. Should be very obvious if they're pooched with the solid subframe/diff bushings.

While the diff was off, I figured I was already halfway to having access to the brake hoses above the subframe, so I swapped the original hoses for the BMA Stainless hoses I bought a few years back. Absolute cock of a job, no idea why BMW didn't just put them like 0.5" further forward...


Tuning escapades:

Car looked like it was running like 1 pt richer than last year, but it was just the LC-1 losing calibration. Leaning toward replacement this/next year with something newer.

Trying to resolve a few weird issues:

Car might have weird TPS adaptations (12.53 - 13.06% is where it seems to stabilize - I have seen this range of adapts in logs from others though?)
Car idles ~2000 RPM on start unless I tap the gas pedal, or TPS adapts are reset.
Car goes quite lean on tip-in. - I've tried to add fuel to the main fuel table as a first step.

Possible causes:
M50 Throttle body instead of M52 (maybe TPS is angled differently on the shaft?) - Currently cleaning up an M52 TB and have an OBD2 throttle cable on order to test.
S52 Enrichment tables instead of M52, since they are quite different. - 100% an oversight, probably from when I switched to the S52 software years back.

My main theory about this is that something has to work because turbo/maf/modded software aside, nothing has changed off boost.

TPS Adaptions seem okay to me, I recall mine being around that.

I ran M50 throttle body with M50 manifold without any issues, car idled fine.

[Northern]

I only noticed the s52/m52 table snafu while I was writing that post, so I need to take a bit and try to see what the end values look like.

I went for a quick drive around the block on the new diff this evening. I had the windows up and everything off and I think it made more gear noise than my old diff? (solid bushings in the rear)
I did a quick drive to drop off some tools and with the windows down, I don't hear anything so maybe I'm just being paranoid.

I must be losing it because somehow the new taller diff feels sportier than the short gearing before?
I can't really explain it, but the gears don't feel long and I expected more of a change - maybe to regret it because I thought the 2.93 could be too tall?

In any case, I really like it, but I think my axles really need to be replaced. They have a lot of play and one started to make a clunk/ping noise.
Who knows, I might get a job soon and be able to afford to do things in other than the most jank ways lol

[ba114]

i remember my tps adapts being between 12 and 14%